Cytomegaloviruses (CMVs) are species specific beta herpesviruses that establish life long infection in the majority of mammals. This lifelong residence is provokes very large CD8 and CD4 T cell responses in the majority of hosts, that have a characteristic "effector memory" (TEM) phenotype and tend to increase in numbers with age, sometimes dominating the entire CDST cell compartment in the elderly. Two paradoxes underlie this signature response. First, the majority of CDST cells have a poor ability to recognize infected cells, due to the function of CMVs genes that interfere with antigen presentation. How are these T cells primed? Second, although latent virus is present in infected hosts, there is little evidence of virus activity. Where does the antigen come from that drives the response? These questions will be addressed using the C57BL/6 mouse model of CMV infection. The first specific aim of this proposal is to test the hypothesis that cross-presentation is responsible for priming that majority of the CDST cell response. To address this aim, a unique set of tools is being assembled: the immunodominance hierarchy amongst 25 CMV epitopes recognized by CDS T cells from C57BL/6 mice will be measured. Mutant viruses lacking both the ability to impair antigen presentation and the ability to impair the expression of co-stimulatory molecules on dendritic cells are being generated, and a mouse model of impaired cross presentation will be used. There is a marked alteration in the main antigens recognized between acute and chronic infection. The second aim addresses the cause of that shift. The ability of the immune system to impact viral activity in different cell types, particularly endothelial cells, will be addressed as a primary cause of this altered immunodominant. Finally, the paradox that the large, inflating T cell response is maintained by viral activity that is below the threshold of detection will be addressed. The TEM phenotype of the CDST cells will be evaluated. Are these cells equivalent to the transitory TEM population observed following resolution of acute infection, resulting from constant restimulation with antigen? Or do they represent a quiescent, long-lived population, as has been proposed for human TEM of a similar phenotype?